How many answers were correct
Each bar is one benchmark: how many items it answered correctly out of every item in that set. We wrote the questions and answer keys ourselves and had them checked by two native speakers, and every model answers exactly the same items.
9 benchmarks
Pemahaman slang Indonesia32/42 76.2%
Koreksi ejaan Indonesia22/38 57.9%
Normalisasi teks Indonesiaquestion version 2KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.27/49 55.1%
Normalisasi teks Indonesiaquestion version 1KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.26/49 53.1%
Normalisasi teks Indonesiaquestion version 3KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.23/49 46.9%
Tingkat tutur bahasa Jawamultiple choice11/27 40.7%
Arti kata Jawa sehari-hari25/135 18.5%
Arti kata Jawa langka1/43 2.3%
Tingkat tutur bahasa Jawaopen answer0/6 0.0%
How far the cost pays off
Sorted by the cheapest correct answer first. The left column shows how many items were answered correctly, so a top row with a long left bar means both cheap and accurate.
8 with cost on record
Correct answersCost per correct answer
Normalisasi teks Indonesiaquestion version 1KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.53.1%$0.000014
Normalisasi teks Indonesiaquestion version 3KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.46.9%$0.000016
Normalisasi teks Indonesiaquestion version 2KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.55.1%$0.000017
Tokens used
Input and output are kept apart because they are priced differently, often tenfold. A long output bar means the model talks a lot, and that is where nearly all the cost difference comes from.
97,593 in, 6,663 out
inputoutput
Pemahaman slang Indonesia11,084 · 120
Koreksi ejaan Indonesia11,546 · 221
Normalisasi teks Indonesiaquestion version 2KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.11,349 · 875
Normalisasi teks Indonesiaquestion version 1KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.9,348 · 873
Normalisasi teks Indonesiaquestion version 3KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.10,521 · 879
Tingkat tutur bahasa Jawamultiple choice5,016 · 78
Arti kata Jawa sehari-hari37,698 · 3,509
Tingkat tutur bahasa Jawaopen answer1,031 · 108
1 benchmarks are not drawn here: those runs came from an older path whose ledger did not store token counts. Not recorded, not zero.
Wait time per call
Median wait for a single call on each benchmark. Median rather than mean, so one stalled call does not move the number.
9 with wait time on record
Tingkat tutur bahasa Jawamultiple choice0.55 s
Koreksi ejaan Indonesia0.56 s
Normalisasi teks Indonesiaquestion version 2KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.0.65 s
Normalisasi teks Indonesiaquestion version 3KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.0.67 s
Normalisasi teks Indonesiaquestion version 1KeteranganWe asked this benchmark with several different phrasings of the question, on the same items and the same answer key. Scores across phrasings are not directly comparable; open the benchmark page to see them side by side.0.96 s
Tingkat tutur bahasa Jawaopen answer1.10 s
Arti kata Jawa langka1.40 s
Test coverage
6/ 6
This model has been tested on all 6 benchmarks we publish.
Earlier question versions
The questions or the key changed after these were measured. Kept as history, not as current results.show 2 rowsclose
Pemahaman slang Indonesia22/27 81.5%
Arti kata Jawa sehari-hari21/135 15.6%
Log11 calls to this model, spread across 4 days.OpenClose
This model is called through a provider id that may not pin a version, so the dates below state which weights were actually tested. Rows marked with a hyphen have a date we inferred from file history rather than recorded at call time, so we do not show a clock time for them.
1 runs
- 22:26Pemahaman slang Indonesiamultiple choice32/42
1 runs
- 20:12Arti kata Jawa sehari-hariword meaning25/135
3 runs
- -Keteranganestimated dateArti kata Jawa langkaword meaning1/43
- -Keteranganestimated dateArti kata Jawa sehari-hariword meaning21/135
- 06:03Tingkat tutur bahasa Jawaopen answer0/6
6 runs
- 22:41Tingkat tutur bahasa Jawamultiple choice11/27
- 18:04Normalisasi teks Indonesiaopen answerPrompt stating the key's rule23/49
- 15:58Normalisasi teks Indonesiaopen answerDetailed prompt forbidding word swaps27/49
- 15:02Normalisasi teks Indonesiaopen answerBuilt-in prompt26/49
- 14:55Koreksi ejaan Indonesiaopen answer22/38
- 14:46Pemahaman slang Indonesiamultiple choice22/27